Research On Tank Engine Health Detection And Fault Diagnosis System

Engine,as the power source of tank,is one of the core equipments of tank vehicle.Its running state and performance will directly affect the operational effectiveness of tank.According to the literature and survey statistics,since the first World War,the failure rate caused by mechanical failure in the battle and training of tanks occupies the first place(35%),and the tank engine fault in the mechanical fault accounts for 48%.As the information level of our military equipment is getting higher and higher,as the representative of information equipment,the structure of tanks is becoming more and more complex,and the requirements for equipment support are becoming higher and higher.In order to find out the fault signs of the tank engine in time and eliminate the faults in time to avoid the occurrence of major accidents,it is very necessary and meaningful to monitor and diagnose the tank engine health according to the actual running condition of the tank.This paper takes a certain type of water-cooled turbocharged tank diesel engine as the research object,and this kind of common failure mode of tank engine type based,acquisition,processing and analysis of the key characteristic parameters of tank engine,using statistical methods,clustering algorithm to realize health monitoring running state of tank engine and anomaly detection;Through the research on the fault diagnosis method of Tank Engine Based on the integrated classifier,the classification recognition of the typical fault mode of the tank engine is realized.Firstly,the mechanical system of the tank engine components are briefly introduced,and then analyzed the tank engine fault mechanism and characteristics and summarizes its main failure mode and feature selection.On the basis of the characteristic of the tank,designed the overall architecture,the tank engine health monitoring and fault diagnosis system of the system function module and work flow.The detailed analysis and design of the tank engine data acquisition and transmission management system,the establishment of tank engine vehicle information acquisition platform,wireless communication system based on GPRS and the tank engine monitoring and diagnosis database management system has laid a solid foundation for the data of health monitoring and fault diagnosis of tank engine.Finally,data mining method based on the running state of the tank engine monitoring,in order to improve the accuracy of state monitoring of tank engine,was established based on the statistical method of Tank Engine important state parameter monitoring method and anomaly detection model based on machine learning.Considering the fault condition of Tank Engine category is often difficult to accurately grasp,using the new machine learning algorithm and construct the integrated model,analyze the operation data of Tank Engine on the collected vehicle test system,realize the accurate classification of fault pattern recognition of a variety of tank engine.